List your company
ResearchFormerly WizWand

VLM Reasoning on BlackJack (In-Distribution)

43.8Success RateGFlowVLM w/ DB
5.3215.3125.335.29Mar 9, 2025
Updated 8mo ago

Evaluation results

MethodLinks
GFlowVLM w/ DBTrain Data=Off-Policy, Assump.=Non-Markovian, SFT Init.=Yes2025.0343.8
GFlowVLM w/ Var-TBTrain Data=Off-Policy, Assump.=Non-Markovian, SFT Init.=Yes2025.0343
GFlowVLM w/ SubTBTrain Data=Off-Policy, Assump.=Non-Markovian, SFT Init.=Yes2025.0342.4
GFlowVLM w/ DBTrain Data=On-Policy, Assump.=Non-Markovian, SFT Init.=Yes2025.0342.2
GFlowVLM w/ SubTBTrain Data=On-Policy, Assump.=Non-Markovian, SFT Init.=Yes2025.0341.7
GFlowVLM w/ Var-TBTrain Data=On-Policy, Assump.=Non-Markovian, SFT Init.=Yes2025.0341.4
RL4VLM+Train Data=On-Policy, Assump.=Non-Markovian, SFT Init.=Yes2025.0341
RL4VLMTrain Data=On-Policy, Assump.=Markovian, SFT Init.=Yes2025.0340.2
RL4VLM*Train Data=On-Policy, Assump.=Markovian, SFT Init.=Yes, Reward function=Same as ours2025.0323.5
SFT-w/o- [DONE]Train Data=Off-Policy, Assump.=N/A, SFT Init.=N/A2025.0323.1
SFT-w/- [DONE]Train Data=Off-Policy, Assump.=N/A, SFT Init.=N/A2025.0320.2
GFlowVLM w/ SubTBTrain Data=Off-Policy, Assump.=Non-Markovian, SFT Init.=No2025.0317.4
GFlowVLM w/ DBTrain Data=Off-Policy, Assump.=Non-Markovian, SFT Init.=No2025.0313.8
GFlowVLM w/ SubTBTrain Data=On-Policy, Assump.=Non-Markovian, SFT Init.=No2025.038.4
GFlowVLM w/ DBTrain Data=On-Policy, Assump.=Non-Markovian, SFT Init.=No2025.036.8